Roberto Franzosi (Emory University, Stefania Vicari (University of Leicester)

Transcription

1 What s in a Text? Answers from Frame Analysis and Rhetoric for Measuring Meaning. Roberto Franzosi (Emory University, rfranzo@emory.edu) Stefania Vicari (University of Leicester) Preliminary draft. 2 tables still under construction. 1. Frame Analysis: A Social Science Approach to Text Media Frames Collective Action Frames Frame Analysis, Persuasion, and Rhetoric On Rhetoric Means of Persuasion Anything of Use Here to Frame Analysts? The Five Canons of Rhetoric Invention: The topics Anything of Use Here to Frame Analysts? Arrangement Anything of Use Here to Frame Analysts? Style: Rhetorical Figures Anything of Use Here to Frame Analysts? Frame Analysis and Rhetoric: A Missed Opportunity for a Fruitful Encounter? Bibliography... 37

2 1. Frame Analysis: A Social Science Approach to Text When in 1972 Bateson wrote Steps to an Ecology of Mind, he probably would not have predicted that his work would become central across different social science disciplines for the development of frame analysis. We owe to Bateson the first conception of frame as the set of interpretative frameworks that individuals apply to understand other people s actions and words: past experience is internally categorized and used to face new events and situations. When Goffman (1974) borrowed Bateson s interpretation of frames as mental constructs, he introduced the concept into sociology from anthropology. For Goffman, frames are schemata of interpretation (Goffman, 1974:21), found particularly in texts and with an ongoing dialogue between texts and mental processes; frame analysis aims to investigate processes of signification by looking at the way meanings become functional to organize social experience. From these early beginnings, different disciplines, from psychology to artificial intelligence, communication and media studies, linguistics, political science, anthropology, and sociology, have produced different frame approaches. 1 In this paper, we focus on frame analysis in the fields of communication and media studies and sociology (in particular, in social movement research). We trace both theoretical and methodological developments. We detail the longer and longer list of what to look for in a text that frame analysts came up with as they grappled with issues of measurement of frames in empirical investigations. We then show how twenty five hundred years of rhetoric would have provided frame analysts with a ready-made, and more comprehensive, list. With knowledge of rhetoric lost in the 20 th century, frame analysts simply reinvented the wheel (as it often happens in the production of knowledge). A missed opportunity? 2

3 1.1. Media Frames The idea that media provide audiences constructed versions of reality has been central to communication, media, and cultural studies. In Making News, one of the most cited books in the field, Tuchman (1978: 1; ix, 180; emphasis added) provides an early use of the word frame: News is a window on the world. Through its frame, Americans learn of themselves and others ; the media set the frame in which citizens discuss public events ; news imposes a frame for defining and constructing social reality. It is with Gitlin, however, in another extremely popular book, The Whole World is Watching (1980), that the notion of frame was to become central: What makes the world beyond direct experience look natural is a media frame. (1980:6; original emphasis) Media frames are then structures of cognition and interpretation (Gitlin, 1980: 22), structures that change over time and become the taken-forgranted conventional wisdom, the hegemonic definitions of how things are (Gitlin, 1980: 303). To frame, Entman would later write, in a definition that was to stick, is to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described (1993: 52). Frames, then, define problems, diagnose causes, make moral judgements, and suggest solutions. It is one thing is to provide concepts and definitions and another to measure them.tuchman and Gitlin take a qualitative approach to measuring. Tuchman (1978), although she does note some of the linguistic characteristics of news (1978:106) short paragraphs and sentences, insistence upon facts, news as stories (built around the who, what, when, where, why and how, 1978:134), story line in the past tense and headline in the present she is mostly interested in the broader aspects of framing, framing as the result of media organizational need 3

4 for routinizing of the unexpected : typification of news, location of news bureaus (e.g., in Washington, DC, in Wall Street), professionalization of journalists, web of facticity (i.e., the focus on events rather than issues. Gitlin similarly applies a qualitative, literary approach to news media with the aim of teasing out those determining but hidden assumptions which in their unique ordering remain opaque to quantitative content analysis (Gitlin 1980:303). But contrary to Tuchman, Gitlin focuses on media content, rather than media organizations. Chapter after chapter, news clip after news clip, Gitlin details the framing devices used by the New York Times and CBS News to describe the SDS movement of the 1960s: from early trivialization, polarization, emphasis on internal dissention, marginalization, disparagement by numbers, and disparagement of the movement s effectiveness to later reliance on statements by government officials and other authorities; emphasis on the presence of Communists; emphasis on the carrying of Viet Cong flags; emphasis on violence in demonstrations; delegitimizing use of quotation marks considerable attention to right-wing opposition to the movement (1980: 27-28). Entman, in his seminal review article on framing, recommends quantitative content analysis informed by a theory of framing as a way to identify and describe frames (1993: 57). By the time of Entman s recommendation, Gamson had been toying for over a decade with content analysis as a way to measure frames. It is no doubt to Gamson that we must turn for the earliest attempts to measure frames quantitatively. Admittedly, his 1983[1980] 2 chapter is only a first step in the analysis of the issue culture of social welfare policy under Nixon (Gamson and Lasch 1983:402). But it does take the first steps in that direction: provide a set of stringent definitions of concepts, arrange these concepts in order of decreasing abstraction (a signature matrix ) 3, and propose the categories of the signature matrix as the coding scheme for a content 4

5 analysis of media material (Gamson and Lasch 1983:401). For the content analysis itself we have to wait a few more years (Gamson and Modigliani 1987, 1989). But Gamson and Modigliani s analyses of issues of affirmative action (1987) and nuclear power (1989) do use content analysis on a systematic sample of media content, on the basis of the conceptual apparatus of the signature matrix as coding scheme. 4 And while relying on quantitative content analysis as their main methodological tool, Gamson and Modigliani (1989:11) also provide qualitative analyses especially of visual imagery... Here we attempt to present enough rich textual material so that readers can form their own independent judgments on the validity of our argument. In an unpublished paper presented at the 1991 annual convention of the Association for Education in Journalism and Mass Communication, Tankard et al. similarly attempt to bridge the gap between a quantitative approach and a qualitative approach to the study of news. (1991:1-2) Tankard et al. squarely focus on issues of measurement, and apply standard principles of content analysis (random sampling of articles, coding scheme made up of mutually exclusive categories created inductively, coders training and instructions, inter-coder reliability, quantification by counting occurrences of categories) to the quantitative measurement of a slippery concept. They provide a list of framing mechanisms based on 11 items headlines and kickers, subheads, photographs, photo captions, leads, selection of sources/affiliations, selection of quotes, pull quotes, logos, statistics/charts and graphs, and concluding statements and list a set of indicators based on specific language and arguments [that] serve as indicators for each frame. (1991:7) Tankard et al. tell us that these indicators are constructed ad hoc ( inductively ) and that they are specific to areas or domains of news content but, unfortunately, do not give any indication of how they should be constructed. 5

6 Building on Gamson s work, Pan and Kosicki (1993:58-62) classified framing devices present in news discourse into four structures: syntactical, script, thematic, and rhetorical. Syntactical structures refer to the patterns of arrangement of words and phrases into sentences. Scripts refer to the narrative elements of a text, the familiar five Ws and one H in news writing: who, what, when where, why and how, a structure also known as story grammar (Pan and Kosicki, 1993:60). Thematic structures define how an issue, a theme, rather than actors and actions (a story), is discussed through hypothesis-testing elements (e.g., quotations, journalists reports). 5 Finally, [r]hetorical structures of news discourse describe the stylistic choices made by journalists in relation to their intended effects. (Pan and Kosicki 1993:61) 6. Subsequent empirical media and communication research based on framing has relied on Gamson s and Pan and Kosicki s work for operationalization (e.g., Nelson et al. 1997; Van Gorp 2005, 2007). Tankard (2001) identifies two further approaches to frames beyond media packages : framing as a multidimensional concept the different content-related dimensions of a story that may vary from story to story and the list of frames approach the different aspects of an issue, as underscored by framing devices and framing mechanisms, or focal points for identifying framing, i.e., headlines and kickers, subheads, photographs, photo captions, leads, selection of sources or affiliations, selection of quotes, pull quotes, logos, statistics, charts, and graphs, and concluding sections (Tankard, 2001:101). Again, framing mechanisms do not seem to add much new to Pan and Kosicki s structural dimensions of discourse (i.e., syntactical, script, thematic, and rhetorical) (for a convenient summary of the main concepts and framing devices in the media literature, see Table 1). TABLE 1 ABOUT HERE 6

7 1.2. Collective Action Frames Gamson s work on media and social movements was seminal in the development of both media and collective action frames. But it was Benford and Snow (Snow et al. 1986; Snow and Benford 1988, 1992; Benford and Snow 2000) who provided the main theorization of collective action frames, understood as action-oriented sets of beliefs and meanings that inspire and legitimate the activities and campaigns of a social movement organization (Benford and Snow 2000:614). The final result of this conceptual elaboration is a complex system of nested categories, graphically summarized in Figure 1 for visual convenience. 7 FIGURE 1 ABOUT HERE Collective action frames consist of both characteristic and variable features (Snow et al. 1986:467-76; Snow and Benford 1992:136-41; Benford and Snow 2000:614-22, ). The characteristic features of collective action frames comprise three core framing tasks diagnosis, prognosis, motivation and discursive processes. Core framing tasks provide social movements a shared understanding of some problematic condition or situation they define as in need of change, make attributions regarding who or what is to blame [diagnosis], articulate an alternative set of arrangements [prognosis] and urge others to act in concert to affect change [motivation]. 8 (Benford and Snow 2000:615) Variable features concern those aspects of social movement frames that vary from movement to movement, from frame to frame: problem identification and direction/locus of attribution (also, issues of interest), flexibility and rigidity, inclusivity and exclusivity, interpretive scope and influence, and resonance. Problem identification and direction/locus of attribution 9 refers to the problems or issues a frame focuses upon, flexibility and rigidity, inclusivity and exclusivity measures the degree to which frames are flexible and inclusive, 7

8 capable of embracing different issues (Benford and Snow 2000:618), and the related concept of interpretive scope and influence the degree to which frames are broad ( master frames Benford and Snow 2000:618). Resonance deals with the question: how does a frame resonate with the members of a group it is attempting to mobilize? Does a frame strike shared inner chords or is it at odds with a target audience s values and beliefs? (Snow and Benford 1988: :140) Resonance is made up of credibility and salience. In turn, credibility depends upon a frame s consistency, empirical credibility, and the credibility of frame articulators. Consistency refers to the fit between a Social Movement Organization s (SMO) beliefs and claims, and its actions (Benford and Snow 2000:620). Consistent frames show no contradiction between an SMO s beliefs and actions, words and deeds (Benford and Snow 2000:620) Empirical credibility refers to the extent to which frames show no contradiction between its claims and events in the world (Benford and Snow 2000:620). Finally, the credibility of frame articulators depends upon the degree to which the proponents of a collective action frame are credible in the eyes of their target audience (e.g., in terms of their status and knowledge). A frame s salience to targets of mobilization depends on its centrality, experiential commensurability, and narrative fidelity. Centrality addresses the question: are the beliefs and values represented in a frame crucial to the lives of the intended target audience? Experiential commensurability addresses a related question: is the frame consistent with the personal, everyday experiences of the targets? Finally, narrative fidelity refers to the extent to which frames resonate with the targets broad cultural narrations ( cultural resonance ) (Benford and Snow 2000:621-2). Three sets of overlapping processes contribute to collective action frames: discursive, strategic, and contested (Benford and Snow 2000:623). Discursive processes, which are part of 8

9 frame characteristic features, refer to the talk and conversations and written communications of movement members and consist of frame articulation and frame amplification (Benford and Snow 2000:623). 10 Frame articulation involves the connection and alignment of events and experiences so that they hang together in a relatively unified and compelling fashion. (Benford and Snow 2000:623). Frame amplification (or punctuation 11 ) refers to the foregrounding and backgrounding of specific issues, events, and beliefs (Benford and Snow 2000:623). 12 Strategic or alignment processes are deliberative, utilitarian, and goal directed: frames are developed to achieve a specific purposes to recruit new members, to mobilize adherents, to acquire resources (Benford and Snow 2000:623). In their study on micromobilization, Snow et al. (1986) identify four main strategic alignment efforts: frame bridging, frame amplification, frame extension, and frame transformation. 13 Frame bridging refers to the linking of two or more ideologically congruent but structurally unconnected frames regarding a particular issue or problem. (Benford and Snow 2000:624) Frame amplification involves the idealization, embellishment, clarification, or invigoration of existing values or beliefs. 14 (Benford and Snow 2000:624) Frame extension refers to a frame s attempt to extend beyond its primary interests to include issues and concerns deemed to be dear to its intended target audience (Benford and Snow 2000:624) Finally, frame transformation refers to an SMO s involvement in changing old meanings and/or creating new ones (Benford and Snow 2000:624). 15 The third framing process, contested process, deals with the contested nature of any construction of reality, whether internal or external to a movement (Benford and Snow 2000:625-7). The contested process is made up of counterframing, the setting up of alternative definitions and representations of reality (Zuo and Benford 1995:139), frame disputes/contests, the conflict between frames and counterframes, between a movement s definitions of reality and that of its opponents (Benford 1993), and the 9

10 dialectic between frames and events, the complex interaction between events and frames, events and ideology. By the early 1990s, this rich theoretical development on collective action frames was slowing down. Calls for more empirical work and applications of the concepts started multiplying (e.g., Gerhards and Rucht 1992:573; Benford 1997:411; Snow 2004:386). An empirical approach to frames raised two questions: 1. In which loci do social movements concretely express frames? 2. How can scholars recognize frames and their various features in these loci? The first question led to texts: speeches, pamphlets, radio and TV talks, media news, interviews. And once in the realm of texts, in dealing with the second question, frame analysts found themselves back to Gitlin s and Gamson s work, back to symbolic devices. But they also proposed new things. For instance, Gerhards and Rucht claim that frames should be studied in terms of argumentative structures and thematic components (1992:574). Johnston turned to micro-discourse analysis for help, to such textual elements as the social role of the actor producing the text, non-verbal cues of oral texts, interactional elements emerging in dialogical communication exchanges, and cross-references within the text (Johnston 1995:219). Like Pan and Kosicki before him (1993:60), Johnston also proposed story grammars as a means to uncover the structural elements of a frame in texts, particularly narrative texts (Johnston 1995:235-6, 2002:82). Qualitative scholars have typically approached the measurement of frames by providing snippets of texts, selected as examples of specific frames. That is true even in cutting-edge empirical studies where frames occupy a central role in a paper s explanatory model (e.g., Babb 1996; Diani 1996). It is also true in sophisticated quantitative papers that rely on content analysis to quantify features of a text while providing snippets of texts as frame exemplars (e.g., Cress 10

11 and Snow 2000; McCammon et al. 2004). Yet, when content analysis is used in papers that pay a great deal of attention to methodological issues 16, the coding scheme is never published, so we do not know what exactly was measured (e.g., McCammon et al. 2004; McCammon et al. 2007; Snow et al., 2007) (for a convenient summary of frame primary concepts and of symbolic devices according to the social movement literature, see Table 2). TABLE 2 ABOUT HERE 2. Frame Analysis, Persuasion, and Rhetoric Much of what frame analysts do in their dealings with texts has to do with persuasion: whether to provide audiences with ready-made filters of reality or to win over public opinion and militants to a social movement s cause. For twenty five hundred years the study of persuasion has been the purview of rhetoric, rhetoric as the ars bene dicendi the art (or science) of effective speaking/writing (Quintilian, Inst. Or ) 17. And the purpose of effective speaking is persuasion, as Socrates tells Gorgias: rhetoric is a producer of persuasion (Plato, Gorgias, 453a). Aristotle, in his Rhetoric, states right at the start of the book (1356a): Let rhetoric be [defined as] an ability, in each [particular] case, to see the available means of persuasion. Quintilian (Inst. Or ) stresses that same point that the function of oratory lies in persuading or in speaking in a way adapted to persuade. Thereafter, the refrain that persuasion is the goal of rhetoric became standard On Rhetoric Through the centuries, rhetoric has focused on different aspects of the art of persuasion: from the means of persuasive appeals, to the five canons of rhetoric, the functional parts of a text (the typical text, in classical times, being the oration, senatorial or judicial, and in medieval times, the 11

12 church sermon), and the stylistic embellishments of rhetoric (tropes and figures or schemes; Lausberg 1998: , ). Let us review next these rhetorical categories Means of Persuasion It was Aristotle who, in Rhetoric, first divided the rhetorical means of persuasion (persuasive appeals) into three kinds (1357a): logos, pathos, and ethos (Kennedy 1999: 85). The modes of persuasion are the only true constituents of the art: everything else is merely accessory (Aristotle, 1354a). Logos encompasses the appeals to reason, the logical arguments set out to prove the speaker s conclusions (or disprove the opponent s). Pathos refers to the appeals to emotion, aroused to predispose favourably an audience to one s conclusions. Ethos establishes the orator s good character as a way to build authority and credibility. Although these three appeals can be analyzed separately, they work together in combination toward persuasive ends (Kennedy, 1999: 85; for easy summaries of classical rhetoric, see Murphy et al., 2003) Anything of Use Here to Frame Analysts? Classical rhetorical means of persuasion would provide the broad framework for understanding some of the features of frame analysis. Thus, the core task of motivation, the motivational call to arms, the rationale for engaging in ameliorative collective action, may find the basis for a call to action in any of Aristotle s three means of persuasion. Frame analysts reasoning devices (in their different forms of thematic devices or argumentative structures; Gamson and Lasch 1983; Gamson and Modigliani 1989; Gerhards and Rucht 1992; Pan and Kosicki 1993) can be thought of as appeals to logos or reason. After all, as Perelman writes in his The Realm of Rhetoric 12

13 (1982:9, 12): The aim of argumentation is not to deduce consequences from given premises; it is rather to increase adherence of the members of an audience to theses that are presented for their consent. [Yet] argumentation does not aim solely at gaining a purely intellectual adherence. Argumentation very often aims at inciting action, or at least, at creating a disposition to act. Nothing could be more true for collective action frames. Ethos would similarly allow us to understand the frame variable feature of the credibility of frame articulators The Five Canons of Rhetoric The rhetorical tradition has handed down a five-fold classification of rhetoric, the so-called five canons of rhetoric: invention, arrangement, style, memory, and delivery. First broadly introduced by Aristotle, 20 the classification was canonized in the first century BC in Rhetorica ad Herennium (I.3) and in Cicero s De Inventione (I.VI.9-VII) and De Oratore (I.XXXI.142) (Herrick 2001:96). Invention (inventio), or argument-creation, comprises what is being said, the discovery of valid or seemingly valid arguments to render one s cause plausible through topics of argumentation (Cicero De Inventione I.VII.9). Arrangement (dispositio) focuses on the functional parts of discourse and their order in sequential units, notably: introduction (exordium), narrative or statement of facts (narratio), outline of what follows (partitio or divisio), main body of the speech containing logical arguments as proof (confirmatio or argumentatio), refutation or counterarguments of one s opponent (refutatio), and conclusion with final plea (peroratio). Style (or elocution) (elocutio), the third canon of rhetoric, addresses how (rather than what, the realm of invention) something is said. Style is not just a text s 13

14 superficial ornamentation but the rational choice of how to express ideas through words. Memory (memoria) refers to the complex art of memory, of committing to memory not only a specific public speech, but also the vast store of knowledge that may be required in public debates. Delivery (pronuntiatio or actio) refers to the way a discourse is presented to the audience through voice and gestures. Since memory and delivery were often excluded from the realm of rhetoric in post-classical treatises of rhetoric, we will focus here on the first three canons: invention, arrangement, and style Invention: The topics Invention Cicero writes in his De inventione (I.VII.9) is the discovery of valid or seemingly valid arguments to render one s cause plausible. Of all aspects of rhetoric, inventio is extremely difficult (Ad Her. II.1) And that discovery, in classical rhetoric, relies on topics. If we want to track down some argument we ought to know the places or topics the region of an argument (Cicero Topica I.II.7-8) (Greek topoi, Latin loci, literally places ; invention from Latin invenire to find). 21 It was Aristotle who first introduced topics in his Rhetoric as the places where to find arguments and divided them into special topics (Book 1, Chapter 3) and common topics (Book II, Chapter 23). 22 Special topics refer to the specific content of an argument. Aristotle discusses three special topics, resulting in the three main species of rhetoric: judicial (or forensic), deliberative (legislative), and epideictic (ceremonial) (Aristotle 1358b-1359a, 1359b- 14

15 1377b). Each species is defined by its subject matter and its relation to time (past, present, future). Deliberative (or political) oratory initially centered on legislative politics. Its primary purpose was to get the audience to engage in action through exhortation and dissuasion. As such, deliberative oratory is concerned with future events that could benefit or harm society. Four topics of invention, grouped in pairs, have been advanced as pertaining to deliberative oratory: good/unworthy and advantageous/harmful (or pleasant/unpleasant). Judicial (or forensic) oratory originally addressed the sphere of forensic argumentation, oriented towards speeches of either accusation or defense. The judicial orator used narratives of past events to defend or accuse an individual in court. Two main topics of invention are commonly associated with this branch of oratory: the just and the unjust (or the right and the wrong). Epideictic oratory is mainly used to praise or blame people or events. It is concerned with present events, e.g., ceremonies or speeches in general. Epideictic oratory uses two specific topics of invention, virtue and vice (or honorable/dishonorable). Common topics refer to lines of argument that apply equally well to all three species of rhetoric. Aristotle deals with common topics in Rhetoric (1397a-1403a). Cicero later dedicates a short book (Topica) to the issue, listing 17 topics broadly classified into: definition, comparison, relationship, circumstance, and testimony. 23 Definition involves the definition of the issue at hand, linking something to the larger group to which it belongs. The subtopics of division (the whole and 15

16 its constituent parts), genus/species (something as part of a larger class or genus), and subject/adjuncts (the essential of something, or subject, and the accidental or adjuncts). Comparison looks at things for their similarity/difference and degree (more and less, inferior and superior). Comparison, of particular importance to Aristotle because it is the basis of metaphor, is a topic closely related to relationship. Relationship builds connections between things, of cause/effect (the effects of a given cause or the causes contributing to given effects, one of the main sources of argument), antecedent/consequence (the relation between former and consequent events), contraries (the relation between opposite elements), contradictions (concerning the truth or falsity of two propositions). Definition and division also establish relationships of a thing to its group and a whole and its constituent parts. Circumstances deal with the subtopics of the possible/impossible and past fact/future fact to establish the probability of something (whether it is possible or impossible, past fact or not, and the combination of these two axes). Testimony comprises as set of subtopics that deal with external sources of bolstering one s argument: authority (depending upon the expertise and character of an individual called upon to uphold one s argument), rumors (the gossips still used today to discredit others), maxims or proverbs (using famous saying or precept), but also the supernatural, laws, precedents, or examples, and oaths, vows, or pledges

17 Anything of Use Here to Frame Analysts? In the topics of invention frame analysts would have truly found a treasure chest of helpful concepts. Nearly all main frame concepts have equivalents in this part of rhetoric. Certainly, the special topics would provide the foundations of the three framing tasks of diagnosis and prognosis. Judicial (or forensic) oratory uses past events to provide argumentations for what is just and unjust (right and wrong), similarly to diagnosis which expresses moral indignation by highlighting unjust conditions. The topics in judicial oratory useful for diagnosis are: incentives for wrongdoings, states of mind of wrongdoers, kind of persons wronged, classification of just and unjust actions, and comparative evaluation of unjust actions (Kennedy 1999:86-87). Deliberative oratory, with its paired topics of invention of good/unworthy and advantageous/ harmful, focuses on future events to ascertain what action would benefit or harm society, similarly to prognosis, the proposed solution to the diagnosed problem. Epideictic rhetoric, the rhetoric of blame (Lat. vituperatio, Gr. psogos) and praise (encomium) for what is honorable or dishonorable could explain the attributional function of frames as this function attributes blame to culpable agents (diagnosis) and moral responsibility for engagement in future collective action (prognosis). Among the common topics, the topic of definition and its subspecies, to the extent that they draw attention to how something is defined (e.g., an issue, an action), could help understand diagnosis and the punctuating function of the frame characteristic features since this function highlights specific societal elements. The common topic of relationship (particularly, cause/effect) can explain some frame characteristic features: diagnosis, to the extent that this involves the attribution of causality (Snow and Benford 1992:138; also Snow and Benford 1988:200) and, together with its subtopics of cause/effect, antecedent/consequence, contraries, and contradictions, articulation (discursive processes), the connection and alignment of events 17

18 and experiences, and bridging (strategic alignment process). It can also such frame variable features as issues of interest and their attributions since this function assigns effects to internal and external causes. Counterframing and frame disputes fundamentally involve the use of such subtopics of relationship as contraries (the relation between opposite elements) and contradictions. Topics, of course, are not mutually exclusive. Definition may involve relationship and comparison, and relationship and comparison often go together (particularly, similarity/difference and degree). That is certainly the case in amplification, one of the central categories of rhetoric. Indeed, for Quintilian, in amplification and attenuation, there lies all the orator s power Inst. Ora ). Crucial to amplification is the use of certain topics (e.g., relation or comparison, ). And amplification covers both res and verba. You can amplify by adding more things (res, literally), more substantive issues, or you can amplify by means of words. In the Renaissance, the Dutch humanist Desiderius Erasmus would reiterate Quintilian s points in a long treatise first published in 1512, with the title De Utraque Verborum ac Rerum Copia (Copia: Foundations of the Abundant Style), with new editions in 1514, 1526, In one of the beginning sections of Copia, Erasmus writes (1978:301): The abundant style quite obviously has two aspects. subject-matter [res] and expression [verba]. Richness of expression involves synonyms, heterosis or enallage, metaphor, variation in word form, equivalence, and other similar methods of diversifying diction. Richness of subject-matter involves the assembling, explaining, and amplifying of arguments by the use of examples, comparisons, similarities, dissimilarities, opposites, and other like procedures Peacham, who was deeply influenced by Erasmus, would similarly write (1593:120): 18

19 And forasmuch as the principall part of Eloquence standeth by increasing and diminishing, distributing and describing, comparing and collecting: I will first shew what amplification is, how it is divided, the use of it, and also what matters and causes are meetest for it, and after I will proceede to the particular treatise of every figure in their severall orders. Amplification consisteth either of words or of things. We find amplification behind some of the key framing concepts, from mobilizing potency to amplification (or elaboration), extension, and scope and influence. To the extent that amplification and its contrary attenuation involve simple operations of addition and subtraction, rhetorical amplification can help explain transformation which entails changing old meanings and/or generation new ones. More generally, the rhetorical tradition proposes four categories of change (Quintilian s quadripartita ratio, Inst. Or ; Lausberg 1998: 462): addition, subtraction, transposition, and substitution. These are rhetorical strategies for the manipulation and variation of discourse at various levels word forms, sentences, paragraphs, entire texts or speeches and across different levels of rhetoric from invention to style. Finally, the topic of testimony, with its various subtopics, would help frame analysts understand some aspects of resonance, one the frame variable features, notably, the credibility of frame articulators and empirical credibility. Narrative fidelity can also be increased through recourse to such external sources as testimony Arrangement The idea that texts are characterized by distinct functional parts that follow each other in specific, prescribed orders goes back to the early days of rhetoric (see Aristotle Rhetoric 1414b). A sixpart division in introduction (exordium), narrative (narratio), partition (the plan of the speech), 19

20 confirmation (or proof, confirmatio), refutation (reprehensio), and peroration (or conclusion, conclusio) was to become standard, as found in De Inventione (I.XIV.19) Cicero s earliest rhetorical treatise, written approximately at the age of twenty in a rigid, pompous, and didactic manner of presentation yet, ironically, to become the standard for excellence through the middle ages (Murphy et al. 2003:158-9). Three of the six parts of a classical oration are of particular interest in relation to the concerns of frame analysis: narrative/statement of facts, proof, and refutation. 25 Narrative is an exposition of events that have occurred or are supposed to have occurred. (De Inv. I.XIX.27; on narration see Quintilian Inst. Or. IV ) One of the rhetorical exercises Aphthonius prescribes is on narrative. Aphthonius, a teacher active at Antioch between the end of the fourth and the beginning of the fifth century BC, left a booklet of exercises (the progymnasmata) that became very popular in European grammar schools throughout the sixteenth and seventeenth centuries in Latin translations frequently reprinted. (Clark 1952:260) He writes: The tale is concerned with six considerations: the personal agent, the thing done, at what time, in what place, in what manner, and for what cause. (Nadeau 1952:265) A narrative should be brief, clear, and plausible. (Cicero De Inv. I.XX.28) A narrative is plausible (probabilis; Quintilian uses the adjective credible, credibilis; IV.2.52; very similis in the Ad Herennium I.14, I.16) when it seems to embody characteristics which are accustomed to appear in real life if the story fits with the nature of the actors in it, the habits of ordinary people and the beliefs of the audience. (De Inv. I.XXI.29) [N]arrative credibility [also] depends upon narrator s authority (Inst. Or. IV.2.125). The purpose of narrative is not simply a statement of facts but persuasion (Inst. Or. IV.2.21, 31). As a result, silence and emphasis must 20

21 govern the choice of narrative facts (Inst. Or. IV.2.77, 83). In Cicero s words: the speaker must bend everything to the advantage of his case, by passing over all things that make against it which can be passed over, by touching lightly on what must be mentioned, and by telling his own side of the story carefully and clearly. (De Inv. I.XXI.30) The narrative must start sowing the seeds of proofs, of the mitigating or aggravating circumstances of a case in a trial that are the proper subject of the next part of an oration: confirmation or proof (Inst. Or. IV.2.54; the theory of circumstances is also known as peristasis; see Robertson 1946:9-10). Confirmation or proof is the part of the oration which by marshaling arguments lends credit, authority, and support to our case. And those arguments pertain to both attributes of persons and of actions. 26 (De Inv. I.XXIV.34) For person, Cicero lists such attributes as name, sex, race, place of birth, family, age, but also height, physical strength and appearance, intelligence, and other personality traits. Among the attributes of action, we find time, space, reason, manner, or outcome. Quintilian, in his Institutio Oratoria closely follows Cicero s treatment in dealing with the credibility of a story (the extent to which it squares with the character of the person and time and place and other characteristics of the action) (Inst. Or ). Of the 10 methods of amplification Erasmus discusses in Copia, Method 8, consists in the use of the circumstances peculiar to the case peristases partly of non-personal details place, instrument, time... partly of personal details: race, country, sex, age p. 591). According to Erasmus, one can perceive the true orator anywhere in the speech by the way details of this sort are aptly added to the mixture in the appropriate place. this feature pervades the whole speech (p. 592) The refutation is that part of an oration in which arguments are used to impair, disprove, or weaken the confirmation or proof in our opponents; speech. (De Inv. I.XLII.78) As Cicero 21

22 tells his reader, refutation relies on the same forms of invention of confirmation because any proposition can be attacked by the same methods of reasoning by which it can be supported. (De Inv. I.XLII.78) Confirmation and refutation are also two of Aphthonius s progymnasmata, in fact, the only two rhetorical exercises, according to Aphthonius, that encompass within themselves all the power of the art. (Nadeau 1952:268, 270) Anything of Use Here to Frame Analysts? At the level of arrangement, frame analysts would have found more ammunition for their conceptual armory. The theory of circumstances, as found in narrative and confirmation, would have given Pan and Kosicki s (1993:60) and Johnston s (1995:235-6, 2002:82) a solid foundation for their recommendation of using story grammars to uncover the structural elements of a frame in texts. After all, the 7 loci of peristasis, as laid out by Aphthonius in his progymnasmata, are nothing but the five Ws and H of story grammars: Who, What, When, Where, How, and Why (on narrative and the history of the five Ws in rhetoric, see Franzosi, 2011). Narrative would similarly help them with the variable feature of resonance. Both aspects of resonance credibility and salience depend upon rhetorical characteristics of narrative (it must be plausible or credible). In particular, empirical credibility depends upon the rhetorical circumstances of the issue. Some of Quintilian s remarks on narrative shed further light on other aspects of frame analysis and their link to narrative. Quintilian s narrator s authority is nothing but the credibility of frame articulators. Similarly, Quintilian s recommendation of the use of narrative silence and emphasis finds a parallel in social movement frames, in the highlighting of issues in both diagnosis and strategic processes (or alignment), where both amplification and transformation require backgrounding and foregrounding of issues. 22

23 Confirmation and refutation, on the other hand, would help shed light on aspects of the contested framing process: counterframing and frame disputes (rhetoric has nothing to say about the dialectic between frames and events, a modern problem linked to the study of media effects). And that refutation can depend upon a range of forms of appeal: logical, emotional, ethical, or by the use of wit or eloquence (Corbett and Connors 1999:278-9). The various sections Anything of Use Here to Frame Analysts? should leave the reader no doubt about the answer to that question. Indeed, frame analysts would have found in rhetoric a plethora of useful concepts. TABLE 3 ABOUT HERE As the snapshot picture of Table 3, on the relationship between framing and rhetorical concepts, shows, all the main concepts of frame analysis find an equivalent in rhetoric. Yet, rhetoric has not exhausted all its possibilities of usefulness to frame analysts. There is more on offer Style: Rhetorical Figures An orator, to be effective in persuading, must accomplish three goals, according to Cicero: instruct his listener [docere], give him pleasure [delectare], stir his emotions [movere] (Cicero, Brutus, 185; De optimo genere oratorum I.3). 27 To these three goals of the orator there correspond three different styles: the plain style for proof, the middle style for pleasure, the vigorous style for persuasion. (Cicero Orator, 69; see also De Or. II.XXVIII and Brutus, 185). 28 The basic ingredients of style have traditionally fallen under the rubric of tropes and schemes (generally subsumed under the term figure in modern rhetoric although in earlier periods the term figure was a synonym for scheme; Lausberg 1998: , ). 23

24 Sherry, in his Treatise of Schemes and Tropes (1550), would write: thys darre I saye, no eloquente wryter maye be perceiued as he shulde be, wythoute the knowledge of them[figures]. Already in classical rhetoric we find a discussion of figures. 29 Both tropes and schemes involve a change in the forme of words, oration, or sentence, made new by art, differing from the vulgar maner and custome of writing or speaking as Peacham writes in his The Garden of Eloquence (1593:1). But while the trope changes the meaning of words or sentences, the scheme (or schemates or the Latin figura) 30 Roman rhetoricians divided the figures into figures of speech (verborum exornationes or figurae verborum) related to verbal expression, and figures of thought (sententiarum exornationes or figurae sententiarum), related to ideas and arguments (and therefore, closely related to topics of invention) (e.g., Rhetorica ad Herennium IV.13 and IV.46; Quintilian Inst. Or ). In Peacham s words (1593:1): A Trope is an artificial alteration of a word, or a sentence, from the proper and natural signification to another not proper, but yet nigh, and likely. 31 Trope or figure, the purpose of their use is the same: to add strength to arguments and give them grace (Quintilian, Inst. Or ). The number of rhetorical figures grew considerably over the centuries, the Renaissance, with its obsession for classification, bringing out an explosion of terms and sub-terms, from the single-minded emphasis on metaphor by Aristotle in his Rhetoric and the handful of early Gorgian figures to the 64 figures that two centuries later appear in Book IV of Rhetorica ad Herennium. 32 Donatus, introducing a distinction in rhetorical figures between tropes and schemate, in the sections De schematibus and De tropis of his Ars Maior states that there are many schemes, but that the most important ones are only 17 and provides definitions for each; similarly, he lists and defines 13 tropes. Later medieval grammarians and rhetoricians were more prolific in their lists. Alexander of Villa-Dei s Doctrinale (1199) treats 80 figures and Evrard of 24

25 Béhune Graecismum (1212) covers 103 figures (Camargo 1983: 106; on the number of figures in medieval rhetorical textbooks, see also Murphy 1974:183, 185, 189, 190). And in the Renaissance that number climbed to 132 figures with Susenbrotus and to nearly 200 with Peacham. 33 Medieval rhetoricians also set in motion a process of increasing specialization on tropes and figures. The last part of the popular 1 st century BC Rhetorica ad Herennium (Book IV dealing with rhetorical figures) started being circulated separately from the rest of the book. Some of the key medieval texts on rhetoric (and grammar) closely followed Book IV of the Rhetorica ad Herennium: from Donatus s Ars Maior (4 th century AD), with its two brief sections De tropis and De schematibus (and these too were being circulated separately from the rest of the Ars Maior) to the Venerable Bede s De schematibus et tropis (written in 701; in Halm ; also partly translated in Copeland and Sluiter, 2009: ). The medievalists interest in rhetorical figures continued well into the Renaissance with several new specialized texts, from Petrus Mosellanus s Tabulae de Schematibus et Tropis (1516) to Johannes Susenbrotus s Epitome troporum ac schematum (1540), Richard Sherry s Treatise of Schemes and Tropes, the first treatise in English vernacular (1550), Henry Peacham s The Garden of Eloquence, Containing the Most Excellent Ornaments, Exornations, Lightes, Flowers, and Forms of Speech, Commonly Called the Figures of Rethorike (1577, 1593), John Hoskins s Directions for Speech and Style (circa 1600). 34 Little by little, this emphasis on style translated into an emphasis on figures, figures on the four master tropes (metaphor, metonymy, synecdoche, irony), and these on metaphor, metaphor taking over the field as the trope of tropes (Sojcher), the figure of figures (Deguy). 35 As Genette put it: [A]t the beginning of the twentieth century, metaphor was one 25

26 of the rare terms to survive the great shipwreck of rhetoric. (1982:114) Thus, twentieth-century social scientists working on frames were in good company when they narrowly and generically focused on metaphors as a means to study frames. And yet, it would be a mistake to think that Renaissance rhetoricians, with their specialized treatment of figures separate from the rest of rhetoric, were unaware of the complex relation between different parts and levels of rhetoric (on this point, see Sister Miriam Joseph 1947:34, 36; Ragsdale 1965; Varga 1983:88-89; Vickers 1988:319). Neither in classical nor in Renaissance rhetoric was there such disconnect. No doubt, both Cicero and Quintilian were well aware of the nature of rhetoric as an organic whole where the various parts and levels are all interrelated. At the end of his Topica and after a series of (rhetorical) questions, Cicero remarks (23.87): Our next task is to consider what topics are suited to each question. 36 He goes on to list the topics most suitable for the different types of speeches (deliberative, judicial, epideictic) (24.91) and for the different parts of speech (e.g., introduction, narrative, conclusion) (26.97). But if that is true for topics, when it comes to the figures function Cicero is disappointing. (Vickers 1970:98; see also Brennan 1960:60) With little interest in taxonomy, Cicero did not go beyond providing a simple list of figures without naming them (e.g., Orator , De Oratore ; despite his cursory treatment of figures Cicero knew his trade and used a large number of rhetorical figures very effectively for his purposes; Vickers 1988:312). We have to wait for the Rhetorica ad Herennium, traditionally wrongly attributed to Cicero, and for Quintilian for a more extensive treatment of the figures. The last book (Book 4) of the ad Herennium is wholly dedicated to the discussion of 64 figures, with definitions, examples of use, and, occasionally, functions and pitfalls (e.g., personification may be applied to a variety of things It is most useful in dealing with amplification and appeal to pity, IV.66). Quintilian 26

27 dedicates two books (8 and 9) of his Institutio Oratoria to figures, defining each figure, providing examples of its use, and, although mostly interested in general principles, 37 occasionally commenting on a figure s function and its relationship to means of persuasion (logos, pathos, ethos) 38 and parts of speech. 39 Renaissance rhetoricians were no less aware of the connection between style and the rest of rhetoric (invention in particular) even when they wrote specialized treatises on figures. Indeed, it would be a mistake to regard the treatises of the figurists as limited to a discussion of style in the narrow sense. (Sister Miriam Joseph 1947:34) Both Melanchton and Peacham, for instance, while classifying figures along the traditional lines (of words and thought, tropes) they also provide separate classifications by means of persuasive appeals (Sister Miriam Joseph 1947:38-39). Thus, in his Elementa rhetorices (1531), Melanchthon (1531:48-54) classified approximately forty of the figures of thought under selected topics of logic (definition, division, cause, contraries, similitudes, genus, circumstances and signs) (Sister Miriam Joseph 1947:38-39). Among 16 th century English rhetoricians, Peacham stands out for his innovative approach to figures. In the second edition (1593) of his The Garden of Eloquence, Peacham 40 divided his 193 figures (Crane 1954:11) into three orders: 41 a first order of grammatical figures ( figures of words ) 42, and a second and third order of figures of sentences 43, with the second order figures being mainly of pathos and the third of logos. 44 (Peacham 1593:120) Peacham does not content himself with depicting the relationship between means of persuasions and figures in broad paint strokes. For each figure, he also provides a brief section titled The use of this figure where he specifies what the figure should be used for (and in a subsequent section titled The Caution he also warns the reader about pitfalls to avoid in the use of the figure). Thus, for pathos, among the figures of exclamation, ecphonesis can be used to 27